Improving silica compound processing: optimization of the mixing equipment.

Numerous projects to improve silica technology have been carried out during the last few years. Most of the work was focused on new coupling agents, particularly silanes, and new silica types, such as highly-dispersible silica. Less attention has been paid to the adjustment or development of the processing equipment, such as special mixers for a more economical and processing-safe mixing of this type of compound.

It is a common practice to mix silica compounds in the traditional black mixing equipment. They are preferably mixed in intermeshing mixers due to the good temperature control (ref. 1), but tangential mixers are used as well. This equipment is designed for dispersing and mixing, but not for a chemical reaction, as required in the case of silica compounds. Therefore. silica compounds require additional steps to complete the reaction between the silanol-groups of the silica and the ethoxy-groups of the coupling agent. This chemical reaction requires temperatures above 130[degrees]C for the initiation of the reaction, but is preferably performed at higher temperatures in order to have a high reaction rate. On the other hand, the silanization temperature should be well below the scorch temperature. A temperature level around 145[degrees]C is commonly used for the silanization step. In order to stay in this relatively narrow temperature window, silica compounds require several mixing stages with maximum temperatures of approximately 150[degrees]C.

Ethanol formation during the coupling reaction between silica and silane is another problem. Each gram of silane used in the compound forms approximately 0.5 gram (0.64 ml) ethanol when all ethoxy-groups have reacted. In a production plant where large batch sizes are processed, this accumulates to a significant amount of alcohol that either re-condenses in the mixing chamber or is evaporated. When ethanol is not removed from the mixer, it causes slippage of the compound, resulting in a less effective mixing. Additionally, if the ethanol concentration in the batch is high, it delays the silanization reaction.

These problems related to processing of silica compounds have mainly been approached from the raw materials side. The newly developed silica types are highly dispersible, reducing the mixing time required for a good dispersion, a prerequisite for a good silanization of the filler particles. These silica types can be dispersed down to the level of the primary particles (ref. 2).

New silanes have been developed with the aim to reduce the scorch risk during mixing and silanization. These coupling agents contain less sulfur, mono (refs. 3 and 4) or disulfides instead of tetrasulfides, e.g., bis-(triethoxysilylpropyl) disulfide (TESPD) instead of bis-(triethoxysilylpropyl)tetrasulfide (TESFF), or even sulfur-free dispersion agents (ref. 5). Other variants of the most commonly used silane. TESPT, have been tested, partly with very good results. The mono-ethoxysilyl-propylsilane, a silane that couples to the silica by one ethoxy-group only, has the advantage that the amount of alcohol generated during the silanization reaction is only one third compared to the amount of alcohol generated by the reaction of TESPT, but the properties of the cured articles are similar (ref. 2).

Some work was done to improve processing of silica compounds, e.g., by adjustments of compounding or by choosing the most appropriate existing mixer for silica compounds. The scorch risk, for example, can be reduced by the addition of zinc oxide in the final mixing step, instead of addition in the first mixing step (ref. 6). In terms of mixing equipment, the intermeshing mixer is preferred because it shows a tight temperature control, allowing it to work close to the scorch temperature (ref. 1). A further improvement can be achieved by working in a tandem-mixer. The first mixer, a normal mixer that is run under standard mixing conditions, is used for mixing and dispersion. A second mixer is especially designed for the silan-ization reaction (ref. 7).

The silanization reaction

During the silanization reaction, the hydrophilic character of the silica filler surface is changed to a more hydrophobic one, increasing the compatibility of the filler with the polymer. The coupling agent, a silane, reacts with the polar silanol groups on the silica surface to form a shell around the filler particle with carbon and sulfur atoms on the outside. This primary reaction starts at temperatures around 130[degrees]C.

The standard coupling agents are bifunctional silanes, and the most common one is bis-(triethoxysilylpropyl)tetrasulfide (TESPT). The first step of the reaction between the coupling agent and the silanol groups on the filler surface is shown in figure 1.

[FIGURE 1 OMITTED]

In the second step of the silanization reaction, the remaining ethoxy groups of the coupling agent react with adjacent ethoxy groups or silanol groups. This chemical reaction between the filler particle and the coupling agent is an equilibrium reaction. Consequently, in order to achieve a good silanization, removal of ethanol is of key importance (ref. 8).

The sulfur contained in a common coupling agent is necessary to build up the filler-polymer network during vulcanization, giving the silica compound its special property profile. At the same time, it is the limiting factor for the silanization reaction; the upper limit of the temperature window during processing is given by the risk of scorch. Therefore, the temperature range for the silanization reaction is limited from approximately 130[degrees]C to 150[degrees]C, depending on the type of coupling agent and silica, the mixing process and the type of mixing equipment (ref. 9).

The Payne effect

The Payne effect is commonly used to quantify the amount of filler-filler interaction in a rubber compound, the latter being inversely related to the hydrophobation of the silica filler by silanization.

It is directly related to the storage modulus G'. The following effects contribute to the storage modulus (refs. 10 and 11):

* The hydrodynamic effect of the filler particles in the polymer matrix described by the Guth, Gold and Smallwood equation (refs. 12 and 13);

* the crosslink network connecting the polymer chains;

* the chemical and physical interactions between the rubber and the filler; and

* the filler-filler interactions.

Only the filler-filler-interactions are strain-dependent on a small strain scale. The filler network is gradually destroyed with increasing strain; the absolute strain values being small compared to the strain values necessary for the breakdown of the polymer-polymer network, well below 100% (ref. 14). This decrease in storage modulus with increasing strain is called the Payne effect (ref. 10). In this investigation, the decrease in storage modulus from 0.56% strain to 100% strain is taken for the evaluation of the silanization reaction, thereby excluding the long-distance forces between the filler and the matrix.

Experimental

To study the relationship between mixer variables and silanization efficiency, one large masterbatch was prepared which was further mixed in different laboratory and production mixers (table 1) under various conditions.

Preparation of the masterbatch

All investigations were done using a passenger car tire tread masterbatch based on a blend of S-SBR and BR with 83.5 phr silica and a silane as coupling agent (TESPT or TESPD). This compound was mixed and pre-dispersed in a 320 liter intermeshing mixer, with the fingerprint of the mixing step given in figure 2. The masterbatch reached the starting temperature for the silanization, 130[degrees]C, approximately 60 seconds before the batch was discharged; the silanization reaction took place to a very limited extent during this period.

[FIGURE 2 OMITTED]

Mixing equipment

Five different mixers, as given in table 1, were used for the investigation of the silanization step. Three intermeshing mixers in different sizes (I5:5.5 liter, I45:45.0 liter, 1320:320.0 liter; I signifying intermeshing) with standard rotor geometry were used in this investigation. Additionally, two tangential mixers of the same scale, but with different rotor geometry, were used. The tangential rotor with full four wing (F4W; T7) geometry has a larger wing in a smaller angle compared to the tangential rotor with ZZ2-geometry (T4), resulting in a high efficiency in terms of radial mixing and dispersion.

The earlier prepared pre-dispersed masterbatches were used for all experiments. This masterbatch was brought to the final silanization temperature in the internal mixer as fast as possible, with the ram closed. After reaching the silanization temperature, this temperature level was held by adjustment of the rotor speed during the silanization

reaction.

Sample preparation and analysis

The degree of silanization was determined by measuring the Mooney viscosity of the compounds directly after the silanization step. These values give a first indication of the dispersion of the filler and the degree of silanization. More accurate results on the silanization reaction are gained by measuring the Payne effect. The decisive difference between the two measurements is the strain at which the measurement is done: The Payne effect is measured only at low strain values (ref. 14); for viscosity measurements, higher strains are applied. The samples were cooled during the period between mixing and measuring in order to avoid re-agglomeration of the filler (ref. 15). The following equipment and conditions have been used for the preparation and analysis of the compound:

* Compound temperature--measured outside the mixer using a thermocouple immediately after discharging of the batch;

* viscosity ML(1+4), 100[degrees]C--Mooney viscometer 2000 E;

* Payne effect (G' at small strains)--rubber process analyzer RPA 2000 (strain sweep: 0.56%-100%, temperature 100[degrees]C, frequency 0.5 Hz);

* evaporated ethanol--adsorption of ethanol on active char, desorption and quantitative determination by gas chromatography; and

* water content in the compound--Karl-Fischer coulometric titration.

Factors investigated

The following factors that were expected to have an influence on the silanization efficiency were investigated:

* Pressureless mixing--the silanization step was done pressureless in an open mixer. The fill factor was optimized for improvement of intake behavior, silanization efficiency and mixer capacity.

* Different mixer and rotor types--mixers with different rotor design and capacities were compared in terms of intake behavior, silanization efficiency and scorch risk.

* Silanization on a mill--silanization on an open mill was expected to be very efficient due to the good ethanol removal.

* Air injection--a very strong current of air was blown through the mixing chamber in order to remove ethanol.

* Mixer temperation--the influence of the temperature settings of the mixing chamber was investigated.

Results

Pressureless mixing

One of the influencing factors of the silanization reaction is the concentration of ethanol in the compound, as well as in the mixer. As the silanization is an equilibrium reaction, lower concentrations of ethanol in the compound were expected to increase the reaction rate. A means to achieve this is to work in an open mixer during the silanization step. In this series of experiments, the pre-dispersed masterbatch, as described in the experimental part, was heated up to the final silanization temperature in a closed mixer. Once the compound reached the silanization temperature (145[degrees]C), the ram was raised and silanization was done pressureless in the open mixer for 150 seconds. The temperature was held constant at 145[degrees]C during the silanization period by adjustment of the rotor speed. After 150 seconds of silanization, the compound was discharged and further analyzed.

Figures 3a and 3b demonstrate the results of the silanization efficiency for different mixers and different fill factors under standard conditions (closed mixer) compared to the pressureless mode (open mixer). The silanization efficiency was measured by viscosity and Payne effect (figures 3a and 3b) respectively).

A significant improvement of the silanization efficiency was found when silanization was done in an open mixer, with a fill factor typical for silica compounds under normal processing conditions (60%), as well as with a reduced optimized fill factor (40%, 45%). This effect is more pronounced in an intermeshing mixer compared to a tangential mixer. The results found on small scale (T4, 15, T7) were verified on a larger scale in the intermeshing 45-liter mixer (145). In the 145 mixer, the increase in silanization efficiency in the pressureless mode was more pronounced compared to the smaller mixer types. The absolute level of silanization was independent of mixer size.

Reduction of the fill factor in order to improve the intake-behavior has an additional positive effect on the silanization efficiency. A combination of both measures, pressureless silanization and reduced fill factor, leads to a further reduction of the Payne effect as well as viscosity. The lowest values for the storage modulus were found for the tangential mixer with the F4W-rotor geometry (T7) and a reduced fill factor in the pressureless mode.

Figure 4a shows the influence of the fill factor on the silanization efficiency for three mixers with different rotor geometries; two tangential rotors with a ZZ2- (T4) and F4W-geometry (T7) and one intermeshing rotor with PES5 geometry (15). The silanization efficiency increases slightly with decreasing fill factor, as can be seen from the Payne effect (figure 4a). This effect was strongest for the intermeshing mixer (15), whereas the tangential mixer with ZZ2-geometry (T4) was only slightly influenced by the fill factor. The level of silanization was comparable for all mixers, in terms of viscosity, as well as Payne effect. The energy consumption, on the other hand, showed significant differences between the mixer types. Figure 4b shows that the specific energy input was lowest for the tangential mixer with F4W-geometry (T7), while energy absorption for the remaining two mixer types was higher, and for both of them on the same level. For the tangential F4W-mixer (T7), the rotor speed necessary to maintain the silanization temperature level depended strongly on the fill factor, but the tangential ZZ2-mixer barely needed an adjustment.

[FIGURE 4 OMITTED]

A comparative investigation was done with three different mixer types (T7, T4, 15). Figure 5a-c shows that during the complete silanization process, the tangential F4W-mixer (T7) gave the best results at all temperature levels investigated (135[degrees]C to 155[degrees]C); under similar processing conditions the viscosity, as well as the Payne effect, were lowest for this rotor type. Especially in the early stages of silanization (low temperature levels, short silanization periods), this mixer performed significantly better. The intermeshing mixer had an intermediate position between the two tangential mixers. When the silanization reaction proceeds (longer silanization period. higher silanization temperature), the difference in Payne effect between the three mixers was diminishing. Parallel to this, the viscosity was reduced, but the values for the different mixers did not converge.

[FIGURE 5 OMITTED]

The specific energy input, given in figure 6a, shows that the compound absorbs more energy in the intermeshing mixer (15) compared to both tangential mixers. When comparing the two tangential mixers, the mixer with F4W-geometry (T7) absorbed a slightly lower amount of energy compared to the tangential mixer with ZZ2-geometry (T4). In terms of total revolutions (figure 6b), the intermeshing mixer (15) took an intermediate position; the tangential mixer with ZZ2-geometry (T4) required the highest number of revolutions for a certain temperature profile, while the tangential mixer with the F4W-rotor (T7) required the lowest number of revolutions.

[FIGURE 6 OMITTED]

Silanization on a mill

The most efficient way to evaporate ethanol generated during the silanization was expected to be a reaction on an open mill. The pre-dispersed compound was brought to silanization temperature in an internal mixer and was transferred to the mill for the silanization step. As a comparison, a different portion of the same batch was put into an internal mixer, heated up and silanized in the same mixer in the pressureless mode. Figure 7 shows the decrease of the Payne effect and viscosity during silanization on the mill in comparison to silanization in the internal mixer. No difference in Payne effect was found between the mixer and the mill, indicating that the silanization efficiency was the same in both cases under these conditions. Only the viscosity decreased faster in the internal mixer during the silanization period of 20 minutes.

Air injection

Another means to improve the removal of ethanol is air injection into the mixer during the silanization step. A first test series was done in the laboratory on a large scale (145). Air was injected from the bottom part of the mixer, and no special outlet for the injected air was conceived. The compound was warmed up to silanization temperature (145[degrees]C) under standard conditions. When the compound reached this temperature, air injection was switched on and the rotor speed was adjusted in order to maintain the silanization temperature. Figure 8a shows the properties of this compound compared to a compound that was silanized under the same conditions, except that air injection was switched off. Air injection lowered the Payne effect, viscosity and water content in the compound, and ethanol removal was more effective. These results were verified by mixing a series of compounds in a 320 liter mixer in a production plant. Again. the silanization was significantly more efficient when air was injected (figure 8b). All other properties were comparable to the properties of a standard silica compound.

Mixer temperature

A method to achieve a better cooling efficiency of the compound is a lower cooling temperature of mixing chamber and rotors. Tests were done in order to measure the effect of a decreasing cooling temperature on the silanization efficiency. Figures 9a and 9b show the effect of this measure in a tangential (figure 9a) and in an intermeshing (figure 9b) mixer. In both cases, viscosity, as well as storage modulus, were reduced when cooling temperature was reduced from 120[degrees]C to 60[degrees]C. At the same time, input of mechanical energy and the total number of revolutions necessary to maintain the temperature level were increased. This effect was stronger in the tangential mixer compared to the intermeshing mixer.

[FIGURE 9 OMITTED]

Discussion and conclusions

Pressureless mixing

High concentrations of ethanol absorbed in the compound limit the efficiency of the silanization reaction. Measures that improve the removal of ethanol thus enhance the silanization reaction. In this investigation, a pressureless silanization step did indeed result in a more effective removal of ethanol and, as a consequence, a lower Payne-effect and viscosity of the compound (figure 3a,b). In this mode, ethanol can evaporate through the hopper opening, and condensation in the mixer with continuous re-absorption of alcohol by the compound is reduced. When silanization is done in an open mixer with a normal fill factor, the intake behavior is irregular, and as a result, the compound is pushed into the hopper where it remains for a few rotor revolutions before it is pulled into the mixing chamber again. The intake behavior can be improved by reduction of the fill factor. It was found that a fill factor of 40% for the intermeshing mixer and 45% for the tangential mixer was the best compromise between good intake behavior on one hand and reduction of the mixer capacity on the other hand.

The fill factor influences the silanization efficiency (figure 4a). The lower the fill factor, the smaller the total amount of ethanol generated in the mixing chamber and the bigger the free volume in the mixer into which ethanol can evaporate. Additionally, the frequency with which a certain portion of the compound passes through the gap between the rotor and the wall is higher with a lower fill factor. The material is sheeted out to a higher extent and evaporation of ethanol out of the compound is enhanced. The influence of the fill factor on the storage modulus was more pronounced in the intermeshing mixer due to the more irregular intake-behavior compared to the tangential mixer. Therefore, the benefit that can be gained by adjustment of the fill factor and improvement of the intake behavior is bigger for the intermeshing mixer.

No significant difference was found between the three mixer configurations in the level of Payneeffect and viscosity. but the trends for the specific energy input differ considerably (figure 4b). For the tangential mixer with ZZ2-geometry (T4), the energy input was only slightly influenced by the fill factor, and the absolute number of revolutions, as a consequence, was only slightly decreased with increasing fill factor. The other extreme was found for the tangential mixer with F4W-geometry (T7); the total number of revolutions during silanization, as well as the energy input, both decreased strongly with increasing fill factor. The overall level of energy absorption was lowest for this mixer type. The mixer with the F4W-rotors was the most efficient one in this series of mixers: energy dissipation to the mixing chamber was low, and shearing forces were high.

When the three mixer types (tangential ZZ2, F4W, intermeshing: PES5) were compared (figures 5a-c and 6a,b), the mixer with the F4W-tangeutial rotor geometry performed best, this in spite of the tact that energy absorption and total number of revolutions were lower. This proves that the F4W-rotor is very efficient in terms of dispersion and creation of fresh surface. The intermeshing rotor with a PESS-geometry was the least efficient one, when operated without a ram, despite a high energy input and a high number of revolutions that should have resulted in the creation of more fresh surface. The reason might be the less regular intake behavior of this mixer type.

The differences in Payneeffect diminished with higher degrees of silanization: whereas the difference in viscosity stayed the same during the whole silanization process. This is the result of two processes: on the one hand, reduction of the viscosity due to the increasing degree of silanization and, on the other hand, a less effective mastication with increasing temperatures and decreasing viscosities.

Silanization on a mill

Silanization on a mill was expected to be very efficient, but experiments have shown that this is not necessarily true (figure 7). In this case, the silanization was done on a mill with a gap size of 4.5 mm. In the internal mixer, the gap between rotor and mixer wall was 3.3 mm. As a result, the batch on the mill was sheeted out thicker, limiting ethanol evaporation out of the compound. Another factor reducing the silanization efficiency is the stock of material on the mill. A certain portion of the material on the mill passes the split less frequently than the compound in the internal mixer. Both factors limit the silanization efficiency on the mill. Under optimized conditions, less material and thinner sheeting on the mill, the efficiency of the silanization on the mill might be further increased.

Air injection

A very efficient way to improve the silanization efficiency is a strong current of air passing through the mixing chamber (figure 8a,b). The injected air drags ethanol out of the mixing chamber, thus improving ethanol evaporation out of the compound and avoiding condensation of alcohol in the mixing chamber. This reduction of ethanol concentration in the compound shifts the balance of the silanization reaction to the product side, resulting in a higher degree of silanization. Another factor is the cooling effect of the air current, making it necessary to mix with a higher rotor speed in order to maintain the temperature level. This results in better mixing and the creation of more fresh surface. This is a very effective measure, easy to implement into an existing mixer.

Mixer temperature control

Temperation of the mixer is a crucial factor for the silanization efficiency. A more effective cooling, e.g., by a lower cooling temperature of the mixer walls, results in a higher input of mechanical energy. This leads to creation of more fresh surface, enhancing evaporation of ethanol. The Payne effect gives evidence of this effect; whereas the viscosity is also influenced by the breakdown of the polymer due to the higher input of mechanical energy (figure 9a,b).

Summary

In this practical investigation, different mixers and mixer adjustments have been tested in the laboratory on different scales, as well as in a production plant, for their influence on the efficiency of the silanization process. The following factors were found to increase the silanization efficiency:

* Mixing pressureless in an open mixer--after the mixing and dispersion step, the following silanization step can be done pressureless in an open mixer. This enhances ethanol evaporation and thus increases the silanization efficiency. The best way to achieve this is working with two mixers: one standard mixer (preferably with intermeshing rotor geometry) for mixing and dispersion, followed by the silanization step done in a specially designed reactor. The silanization reactor should be able to hold the temperature on the desired level and to let ethanol evaporate out of the compound and out of the reaction chamber.

* Rotor design--tangential rotors perform better in terms of intake behavior when working in the pressureless mode, but intermeshing mixers have the advantage of a good temperature control. The F4W-rotors, aggressive tangential rotors with a larger wing and a smaller angle compared to the tangential ZZ2-rotor geometry, turned out to be the most efficient rotors for the silanization in this investigation.

* Air injection has a positive effect on the silanization efficiency without influencing the overall properties of the material.

* Heating of the condensation-sensitive parts (for example the ram) and intensive cooling of the remaining parts increases the silanization efficiency.

Table 1--mixing equipment used in the comparative
investigation of the silanization efficiency

        Volume               Rotor geometry

  T4      3.6 liter        Tangential (ZZ2)
  I5      5.5 liter     Intermeshing (PES5)
  T7      7.6 liter        Tangential (F4W)
 I45     45.0 liter    Intermeshing (PES45)
I320    320.0 liter            Intermeshing

Figure 3a-effect of pressureless mixing on
viscosity for different mixer types and fill factors

ML(1+4), 100[degrees]C, [MU]

      Ram down (standard process)    Ram up (open mixer)

Standard fill factor:
T4         60%
I5         60%

Optimized fill factor:
I45        40%
T7         45%

Silanization temp. 145[degrees]C, time 150 sec.; Cooling temp.
90[degrees]C, 145-40[degrees]C.

Figure 3b-effect of pressureless mixing
on Payne effect different types and
fill factors

G'(0.56%)-G'(100%) [MPa]

      Ram down (standard process)    Ram up (open mixer)

Standard fill factor:
T4         60%
I5         60%

Optimized fill factor:
I45        40%
T7         45%

Silanization temp. 145[degrees]C, time 150 sec.; Cooling temp.
90[degrees]C, 145-40[degrees]C.

Figure 8a--influence of a air injection on
silanization efficiency in a laboratory mixer

G'(0.56%)-G'(100%) (MPa)

ML(1+4), 100[degrees]C (MU)
Evaporated enthanol (mg)

                     ML(1+4), 100    G'(.56%)-G'    Evaporated
                      [degrees]C       (100%)        ethanol

Moisture content compound:
Air injection on                        1.33%
Air injection off                       1.55%

(I45 mixer temp. 40[degrees]C, fill factor 40, silanization temp.
145[degrees]C; silanization time 150 seconds)

References

(1.) D. Berkemeier, W. Haeder and M. Rinker, Rubber World, 34 (July 2001).

(2.) L.A.E.M. Reuvekamp, thesis, University Twente, ISBN 90 365 1856 3 (February, 2003).

(3.) P.G. Joshi, R. W. Cruse, R.J. Pickwell, K.J. Weller, M.H. Hofstetter; E.R. Pohl, M.F. Stout and F.D. Osterholtz, Crompton Corp. literature.

(4.) J. W. ten Brinke, P.J. van Swaaij, L.A.E.M. Reuvekamp and J.W.M. Noordermeer, paper 131A, ACS Rubber Division Meeting, October 2001.

(5.) T. Fruh, L. Steger and L. Heiliger, proceedings of the Kautschuk Herbstkolloquium 2002, Hanover: Germany, 175 (October/November 2002).

(6.) L.A.E.M. Reuvekamp, S.C. Debnath, J.W. ten Brinke. P.J. van Swaaij and J.W.M. Noordermeer; paper 112, ACS Rubber Division Meeting, October 2002.

(7.) J. Peter and H. Sandau (to Continental AG), German patent DE 199 13 047A 1 (23.3.1999).

(8.) H.-D. Luginsland. Degussa-Hills AG, paper presented at the 11th SRC Meeting, Puchov (May 1999).

(9.) L.A.E.M. Reuvekamp, J. W. ten Brinke. P.J. van Swaaij and J.W.M. Noordermeer; University Twente, paper at the Kautschuk Herbstkolloquium, October 2000.

(10.) A.R. Payne, Rubber Chem. Technol., 39, 365 (1966).

(11.) A.R. Payne and R.E. Whittaker, Rubber Chem. Technol. 44, 440 (1971).

(12.) E. Guth and O. Gold, Phys. Rev., 53, 322 (1938).

(13.) H.M. Smallwood, J. Appl. Phys., 15, 758 (1944).

(14.) W. Niedermeier; J. Frohlich and H.-D. Luginsland, Kautschuk Gummi Kunstoffe 55, 7-8 (2002).

(15.) C. Lin et al., Bridgestone/Firestone Research Inc., paper no. 90, ACS Rubber Division Meeting. October 2001.

Wilma Dierkes and Jacques W.M. Noordemeer: University of Twente; Kai-Udo Kelting, Padeborn University; and Andreas Limper: ThyssenKrupp Elastomertechnik